Heretic demos (non-demopack and NM)

Manually measured ExM8 times for the Vile's demos:

 

e1m8 - 0:43 (or maybe high 0:42), final time is then 9:55
e1m8new - probably 0:46, final time is then 8:44
e2m8 - probably 1:40, final time is then 8:13
e3m8 - probably 4:59, final time is then 17:52
e3m8new - probably 5:21, final time is then 13:22
e4m8 - probably 1:27, final time is then 12:41
e5m8 - probably 3:51, final time is then 16:21

 

I wonder what makes Vincent Catalaa's E2 and E3 demos desync for me. I could play back the E1 demo without desync in wHeretic 0.5a9.

Edited by Keyboard_Doomer

Posted (edited)

Share this post


Link to post

Thanks!

Opulent: I wasn't aware of this mediafire zip...seems like I have most of it already but there are definitely some demos I had missed, so thank you for posting this!

Share this post


Link to post

kraflab
It's not difficult for me, no problem, I'm already have this info in my archive. I have long time wanted to playback these Vile's and Bruno's demos, and today I setup ZDoom/Legacy and did it, you somehow pushed me to it, he he
I watched the Vile's demos and used an external stopwatch, have the same results as Keyboard_Doomer for ExM8 maps (+- 1 sec). Bruno's E5M8 not have end episode text in demo and as Searcher already wrote - has 6368 gametics, I think we can safely take this time, then this demo will be 3:01.94.

I also see that now in DSDA may be missing some vanilla Heretic demos, but not very much, just not easy to compare, maybe these demos will be found later.


 

15 hours ago, Keyboard_Doomer said:

I wonder what makes Vincent Catalaa's E2 and E3 demos desync for me. I could play back the E1 demo without desync in wHeretic 0.5a9.

Keyboard_Doomer, I agree with you, because if the engine holds a demo sync well - it should always hold it, right? I do not know the reason, but I can assume that this wHeretic version may have a dependency from the software environment in the system, installed DirectX and Windows versions. I'm use very old PC now, Win9x/Win2000, old video card and DirectX-8 :) For me these demos are played absolutely normally, checked many times. On which system do you run this wHeretic version, if not secret?
Also, now I download these EP2-3 movies from DSDA and checked that I have the same files. Only one link I found at the moment and checked - I'm using exactly this wHeretic version:
https://gamefront.online/files/7263/whtc05a9.zip
In any case, if you have time and interest - we can try to find out the reason, just for the sake of technical interest.


 

Quote

In real time on my system, timed by hand I get: 
h1c1trik.lmp 34 sec 
h1c7trik.lmp 1:54


 

Searcher, it turns out that these demos you play 2 times slower than you need. If you playback in DosBox - maybe not enough your CPU? Try -timedemo in vvHeretic, I think you see that realtics more than gametics twice.
 

Edited by PVS
1 person likes this

Posted (edited)

Share this post


Link to post

Added 18 demos (vile's episodes and some ones from opulent's zip).  I also corrected some information (sm speed runs that were also pacifist) and removed a couple duplicates.

 

By the way, the site has a feed of recently added demos, here: https://dsdarchive.com/feed

 

There were 2 demos by Steve Dudzik in opulent's archive without text files, so I don't know what port they used, and I'm not sure if the times are correct.  They're h3f8-928 and h3m8-635.

 

Thank you everyone for all the help :)

 

I need to start adding these ports to the site as well so that everything needed is mirrored, but that can wait for another day (I think I have copies of all of them, I just need to set it up).

1 person likes this

Share this post


Link to post
8 hours ago, PVS said:

Searcher, it turns out that these demos you play 2 times slower than you need. If you playback in DosBox - maybe not enough your CPU? Try -timedemo in vvHeretic, I think you see that realtics more than gametics twice.

 

I played those back on an old game box with Win XP, in dos box,  it runs time demo at about 69 FPS on that demo,  Roughly twice the normal speed.  At that speed you can watch and still appreciate the run.  :)

Edited by Searcher
Typo

Posted (edited)

Share this post


Link to post

@PVS

 

I'm actually running it through Wine on Linux. I don't know if some incompatibility with DirectX or something else in the system can lead to a desync but my experience with running older source ports as well as original EXEs through Wine/Dosbox is very good and I didn't encounter a problem like this until now. 

 

I'm using the same wHeretic version you linked and I also double-checked there's nothing wrong with the heretic.wad I have so it's truly a mystery to me why I can play back the E1 demo without desync but not E2 and E3.

Share this post


Link to post

Searcher
I expected that you would see the ratio realtics/gametics as I said, oh, I'm fail. In any case, there are no problem with timing of these 2 demos, they contain 565 and 1539 gametics respectively, it will be 0:16 and 0:43 full time with 1-2 sec coop stats screen recorded in demos. Your 0:34 and 1:54 in real time playback is very slow, I do not know exactly what the problem may be for your DosBox, maybe try other DosBox settings for increase the speed, but you know better your system, of course.


Keyboard_Doomer
Wine under Linux this is a serious test for old wHeretic :) I think he may have problems in the easier cases: when wHeretic run on newer Windows versions, newer DirectX versions, etc. It uses old DirectX for render and sound, it can be important, but these are only my assumptions, no more. I think the maximum that you could try - run wHeretic with -nosound -nomusic and 640x480 resolution (maybe even better in 640x400, his default). I do not think that it will help much, only for test, if you have not already tried this way.
 

Edited by PVS

Posted (edited)

Share this post


Link to post

I did play back the demos with -nosound -nomusic because otherwise wHeretic crashes for me. I only need to use the parameters are if I want to play back a demo, though. I can play without using either of them. I also did use 640x480. Using 640x400 or lower results in crash. 640x480 or above runs fine.

Share this post


Link to post

It's interesting, I did not try to play these demos without sound and music, because they work for me normally. Decided to try to play these demos like you, it seems I found something interesting:
 

Spoiler

wHeretic resolution always 640x480
h3sp1104
1. -nosound -nomusic: desync in the beginning E3M3, player does not fall into the yellow door (test: 4 times, always stable)
2. -nosound: crash, not far from the beginning E3M1 (test: 5 times, always stable)
3. -nomusic: OK! Demo play back normally (test: 2 times)

h2sp-844
1. -nosound -nomusic: desync in the beginning E2M8, player enter in the portals area (test: 1 time)
2. -nosound: same desync in the beginning E2M8, player enter in the portals area (test: 1 time)
3. -nomusic: OK! Demo play back normally (test: 1 time)

h1sp-951
1. -nosound -nomusic: OK! Demo play back normally (test: 1 time)
2. -nosound: crash, in the beginning E1M1 (test: 5 times, always stable)
3. -nomusic: crash, in the beginning E1M7 (test: 3 times, always stable)
 


I can reproduce your problem on my system for these 3 demos, for EP2&3 main problem for me this is -nosound. If you can play back EP2&3 only with -nomusic without crashes - I think, demos can play back normally for you. If you do not want to spend time on it - no problem, it is only a technical interest. Can you remember - where did you have desync for EP2&3 when you tested? I'm interested to compare with my desync's and maybe this information be useful to someone else.
 

Share this post


Link to post

Good job, you solved it :)

 

I wouldn't expect -nousound and -nomusic to have an effect on desync but it's clear they do in this case.

I didn't try E2 demo with only -nosound but for all the rest I get the exact same results as you including desyncs and crashes in the same places.

 

Thank you for taking the time to figure this out.

 

Now that I know how to play back the demos without desync I also went on to measure the ExM8 times. I got the same times for E1M8 and E2M8 as you but for E3M8 I measured 1:50 - one second less than the time you got. But that could of course just be an error on my part.

Share this post


Link to post

Great, because now you could see all 3 the first ever recorded Heretic movies demos on the real game engine. However, it's a pity that in 2002 Vincent did not have a choice and he had to use this wHeretic version to record movies, as close to the original Heretic, but not fully compatible with it, vvHeretic1 patch appeared only in 2005.

If you have desyncs and crashes in the same places as me, we can say that wHeretic works stable, because it stable crashes and desyncs on different systems :) I think it sounds funny, but in my opinion - this is a good characteristic of the engine.

Maybe you have any previous wHeretic versions? I began to search for it is too late and until that moment I was able to find only this latest version. I think, that more older wHeretic versions may be closer to vanilla Heretic or fully compatible with it. For example: htuve1m6 by Dave Turner recorded on wHeretic v0.3x build4, have time 16:30 and play back fine for me on Heretic 1.3.

If manually measured demos time and have +- 1 sec as results - this is normal situation, in my opinion. I prefer your result 1:50, it's good if we make Vincent Catalaa faster :)
 

1 person likes this

Share this post


Link to post
On 3 июня 2017 г. at 10:27 PM, kraflab said:

Added 18 demos (vile's episodes and some ones from opulent's zip).  I also corrected some information (sm speed runs that were also pacifist) and removed a couple duplicates.

What kind of demos do you mean? I do not remember some of these situations (SM speed that were also Pacifist), most likely I just did not notice it.

Demos by Steve Dudzik (Lut) from this old topic:
https://www.doomworld.com/vb/doom-general/2457-heretic-e3m8/

I remember that we talked about this demos with Opulent a few years ago. There are 3 demos: SM&BP Speed for E3M8 and BP Max for E5M8.
1. BP Speed 9:29 for E3M8 not Max, because have 2/3 secrets: visually player visits all secrets, but in secret room with ring he used wings and forgets to fall to the floor, in this case - the game engine does not count the secret.
2. BP Speed 9:29 for E3M8 and BP Max 7:02 for E5M8 technically not contain exit: player press "stop record" button near exit. This situation does not contradict the DSDA rules?
3. It is not exactly clear what versions of game-exe's it uses when recording these demos, I'm have information about it only from this old topic. For E3M8 demos Lut indicates Heretic 1.3 (if I understood correctly), but SM Speed 6:35 demo contain end episode text, this is not possible for vanilla Heretic. BP Max 7:02 for E5M8 most likely was recorded on jHeretic 0.93, this is the most logical option for 2001, when these demos were recorded.

All 3 demos normally play back for me on Heretic 1.3. I wrote text files for these demos for my archive, based on that old topic, you can use them if you want, I put them in the attachment.

Another question. It would be nice to add a column with a demo record date in DSDA. I think, only year will be enough. Also, if there is a functionality sort by this date, it will be very easy to find the latest recorded demos.
 

Lut2001.zip

Edited by PVS
1 person likes this

Posted (edited)

Share this post


Link to post

I removed the demos that don't exit, those aren't considered finished runs by dsda standards.  I just put Heretic v1.3 for the engine, since that does play back the demo (since we are not 100% sure what it was recorded with).  That isn't a huge deal to me, as long as someone can look at the archive and knows what is required to play back a given demo.

 

The "new" demos I found in opulent's archive were a variety of non-record runs which (as far as I could tell) weren't stored in the heretic-n archive I downloaded.  Two examples are E3M4 NoMo 1:05 and E3M4s NoMo 0:59 by PVS.

 

I don't remember exactly which demos were marked sm speed but were in fact pacifist (I should have written it down to post here but I forgot to).  Opulent's archive had these corrections, which is why I noticed them at all.  One example is E2M1 in 0:20 by Kimo Xvirus.  The category listed in the text file is sm speed, but in the comments he writes that the run is also pacifist.  This run appears to be the pacifist record.  I'm not sure if it was listed correctly in heretic-n and I can't seem to get that site to load right now so someone can check that later.

 

As far as sorting by recorded date, I can add that without too much trouble.  I need to expand the options for the "recent demos" page at some point anyway, (for instance, filtering by wad, category, etc), I just haven't gotten around to that because it's not as high a priority as some other features.

 

Speaking of new features, I wrote up a little page to compare movies map-by-map, here: https://dsdarchive.com/compare_movies?category=SM+Speed&id=heretic&level=Ep+1

 

It's a bit basic right now because I'm still working on the functionality, but it should work for selecting and comparing different movies.

2 people like this

Share this post


Link to post

I've started uploading ports so that anything needed to play back the demos is available.  When I was reading the text files for the doomsday demos it seemed to imply that old versions of doomsday are equivalent to jheretic?

 

I have demos marked doomsday v1.7.8, jheretic v0.93(se), and jheretic v0.93(se4).

 

What are these "se" versions?  I think pvs mentioned it is his modified port, but I'm not sure where a link is.  Does that port also play back Andrey Budko's doomsday demos?

 

Sorry for all the questions, I just don't have much experience with heretic demos.

Share this post


Link to post

Some of Andrey's demos are marked as Jheretic178, not sure if doomsday will play them back, or if in fact they are the same.  As I recall the thread he released those in had a link to jheretic178 in it.

Share this post


Link to post

Doomsday 1.7.8 still here:
https://sourceforge.net/projects/deng/files/Doomsday Engine/1.7.8/

Doomsday has it own demo format not compatible with original games, within a common package three games with their own versions. On the example of this Doomsday 1.7.8: it includes Doomsday.exe v1.7.8, jHeretic.dll v1.3.4, jHexen.dll and jDoom.dll - I do not remember correct versions at the moment.
Andrey Budko demos play back fine for me, I've watched them many times, for Doomsday of that time there is no strict dependence from the port version, this demos work for me since Doomsday 1.6.0 and ending 1.8.6. But for this Andrey Budko demos I recommend version 1.7.8 and below, for full correct play back.

kraflab, I'm interested for any Doomsday demo records (Heretic, Hexen, Doom), maybe you remember them in the DSDA archive? I have not yet found.

jHeretic 0.93 - beta version from 2000 and this is latest version compatible with vanilla Heretic 1.3 demo format, this version can not play back newer Doomsday demos. "se" this is my small hobby project which I do sometimes, under "se" I mean - special edition, nothing more. For play back this demos not necessary have jHeretic, this is vanilla compatible demos, can be used Heretic 1.3/vvHeretic4/Chocolate Heretic/etc. To date, I do not have 100% compatibility with the original Heretic, a number of restrictions, etc. This version remains the old port for old windows systems, I do not want unnecessary problems for people, for this reason I'm not public this version, this is just a hobby for me.


 

On 9 июня 2017 г. at 0:42 AM, kraflab said:

Speaking of new features, I wrote up a little page to compare movies map-by-map, here: https://dsdarchive.com/compare_movies?category=SM+Speed&id=heretic&level=Ep+1

This is an interesting idea for good visual compare movies, great. If you can do this, then, the dates of demo records for you are not difficult, If you think this column is necessary, of course.

If JCD reads this:
h2s1-020
h2s5-022
h6s1-022

This is also pacifist demos and need to put them in a Pacifist record table, h2s1-022 pacifist too. kraflab, thank you for informing about this.
 

Edited by PVS

Posted (edited)

Share this post


Link to post

OK thanks for the info, I added doomsday 1.7.8 to the site.  As far as other doomsday records I am not aware of any.  I asked shockblast and he pointed out the doom64tc runs, here: http://doomedsda.us/wad2824.html.

 

The thing with demo record dates is that they are just guesses, based on file modification dates.  I extracted them from the .lmp files, but sometimes this information is inaccurate (my guess is that sometimes it was altered when demos were merged with the archives), and sometimes it is unavailable entirely.  Some of that could be confirmed manually of course, based on forum posts and old archive updates.  I don't really want to put that kind of information directly on the table, since I can't be sure of its accuracy, but I will definitely make it available in some way in the future.  There is a graph of demo count by year if you're curious about that in particular, here: https://dsdarchive.com/wads/heretic/stats

1 person likes this

Share this post


Link to post

PVS: I added sorting by record date to the feed, here: https://www.dsdarchive.com/feed?sort_by=record_date

 

There are a bunch of 'Unknown' at the top of the list there because I'm missing dates.  If anyone can easily determine dates for those (from compet-n updates maybe?) I will update them.  Otherwise I'll get to sorting them to the back so they aren't stuck on top the next time I work on the site.

1 person likes this

Share this post


Link to post

@PVS : I'll check if these demo are pacifist, if so I'll update the tables :D

 

@Kraflab :

E2M4 coop SM-speed is 2009-06-27

E5M8 SM Max is 2008-10-12

the rest is 2007-12-23

 

source : http://lsvc.free.fr/Heretic-N/index.html

Share this post


Link to post

In my opinion - with dates it turned out great, I like it. I know that all the strange dates is JCD demos, I do not know why this is happened for him. I noticed this, because in my archive I also use the record dates.

h2c4s028: 2008-10-09
h1p1-030: 2007-12-23
h2p1-024: 2007-12-23
h2p5-024: 2007-12-23
h6p1-026: 2007-12-23
h4p4-033: 2007-12-23
h1n4-059: 2007-12-23
h1n5-056: 2007-12-23
h5s8-336: 2008-10-12

If automatically extracted dates from the .lmp files sometimes there will be such "unknown" results, especially when you start adding Doom demos, there is a huge amount of them, I understand this. I see, you use some logical filter so that the demos fall into "unknown", right?. Another option: if the filter give the "unknown" result -> try automatically extracted dates from the .txt files (if txt present) and check that the dates there are logical. I think, this is will be better than "unknown". For this Heretic demos - date 2007-12-23 this is txt date and JCD himself uses them in Heretic-N tables.

Thanks for the Doom64TC demos link, it's interesting that someone in 2016-2017 record them on this old modification.

EDIT:

While I was writing this - JCD has already specified :)  For h2c4s028 better use his information, of course. My date for this demo from Opulent archive.

Edited by PVS

Posted (edited)

Share this post


Link to post

Thanks, I fixed this on the site.  I agree that the txt file should be used as a fallback since it is better than nothing in cases where the lmp doesn't work.

Share this post


Link to post

I've started working on stats, so I added a measure called the "record index" that attempts to rank people based on something more meaningful than the quantity of demos like on doomedsda.  For instance, jc has much fewer demos than vincent, but he makes a lot of improvements and beats other runs, so his score is higher.  You can find the table here: https://dsdarchive.com/stats

 

I'll probably tweak the formula and maybe add different weightings at some point, but it's interesting to look at anyway.  By the way, I don't want to crowd the thread with dsda news if that's unwanted, so I can make a separate thread if that's anyone's preference.  Right now it just happens that my site exclusively stores heretic runs, so it made sense for me to post here.

Share this post


Link to post

JCD
Your Heretic-N database again unavailable for me, you once again moved to another address? :)

kraflab
I wonder: if I did not beat any records in my demos, then why do I have one point in the DSDA "record index"?

Yes, beat other times it's hard, but maybe add a separate rank, that shows the total contribution of a people? For example, previously Vincent Catalaa used the table of the players total contribution, where he counted the points like this:
- Single level records are awarded by 1 point.
- Multi-level runs (movies): fastest demo is awarded by 5 points, second fastest by 3 points.

If the demo time has been improved, it does not always mean that the old demo is bad, right? This total rank can show well - how much a people put his time, how much they tried, record demos, thus - they has contributed. I think, have this total contribution rank is good, but maybe use this rank only for speed records, not first attempts (FDA), etc.
---------

Also, I will try to show one of my first attempt record, which I talked about some time ago. Now I thought about it and I doubt that this my demos can be of interest to any databases. In first attempts I never try to exit from the map as quickly as possible, rather, on the contrary I'm trying to see the whole level, as far as I can. For this reason - almost always the demos are long-time and closer to the Max demos, maybe more closer to the map overview.

This my example it's something like - average quality demo from me and average length, I also have demos long in an hour/two or more, with 40 deaths, with absolutely stupid situations - when I see the blue key, but can not understand how can I take it, demos where I could not find the exit from the map, demos where I broke the level mechanics and could not continue, etc, etc. This is absolutely true Heretic first attempts, that I recording just for myself.

kraflab, I think that no one have interest and will look at this example, as it usually happens for the Heretic demos, for this reason it would be good if you can find the time and see this demo yourself and tell the truth - need such demos for the DSDA database or not. All the necessary information present in txt, even more.
 

hfaMdn61.zip

1 person likes this

Share this post


Link to post

I think you might have one point because of e3m4s tie with JC, I'm not sure if I decided to give both a point in that case, but that seems possible.  Otherwise it's a bug 😅

Share this post


Link to post

sorry i haven't posted in awhile. i've gotten a few episode movies (sm-speed) but i haven't posted them here, because... well, i don't really have a reason. i've just been kind of lazy. i can't remember which demos i've posted here and which ones i haven't so i'm just going to post them all. some of them i'm not happy with and i'll probably improve again in the future but i figured i should post them here anyways just so kraflab can keep his archives up-to-date.

 

times listed in the zip/demo name are including exm8, just because that's the metric i've been using to compete against myself. time excluding exm8 is included in parentheses

h1sp-654.zip (6:22)

h2sp-649.zip (5:44)

h3sp-828.zip (7:14)

h4sp1055.zip (9:26)

h5sp1413.zip (10:33)

 

here's a youtube playlist with all of these if you don't have a chocolate heretic/vvheretic install on hand

Edited by veovisRC
4 people like this

Share this post


Link to post

In this txt's you're also too lazy to write your nickname completely? In previous EP1,3 you wrote - veovis, in this - veo. For any database is bad, if one person has a different nickname. How it will be correct? :)

Now, record table for SM speed movies - full by veovis! It's hard work, I know it. Periodically I also try record speed movie for EP5, not yet successfully at the moment. Just like an idea: according to the Heretic-N rules - movies can contain a minimum of 2 completed maps, so far no one has recorded a EP6 movie, I wonder why? E6M1 and E6M2 completed maps, E6M3 does not contain exit, then, for first 2 maps - possible record a small EP6 movie! heh... Also, if you record SM speed demos for 5 secret maps - it will be full SM speed Heretic by veovis! Only E6M3 will remain, where possible record only Max with savegame tic in demo.

I watched this demos and noticed one interesting technical detail on E5M7: veovis slightly thicker than JCD! See for yourself - JCD it easily passes through the bars to the blue key, but to veovis needs to help himself with a Phoenix! For me remains a mystery - from where the Heretic engine can know about it? I searched in source code and did not find anything about 'thickness of players' :)

Great records, thanks for uploading!
 

Edited by PVS

Share this post


Link to post

regarding e5m7, i've tried doing it without the phoenix rod and was able to do it once by pure accident, and since then i've never been able to recreate it. so i've just been using the phoenix rod since it's the one i know how to do. maybe i need to lose a few pounds :P

Share this post


Link to post

two demos to post: episode 1 BP-speed movie in 7:34 (without E1M8 = 6:58) and episode 5 SM-speed movie in 12:57 (without E5M8 = 9:59)

h1bs-734.zip

h5sp1257.zip

5 people like this

Share this post


Link to post

episode 2 BP movie in 7:28 (6:09 without m8). that was way harder than i was expecting and now i'm really scared of the other episodes but i want to do them all eventually.

h2bs-728.zip

3 people like this

Share this post


Link to post

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now