Still trying to see if the PIS on the EMU is working Saw it in the front of the 4024 is on. So that’s good Stutter is noticeable
the worst stutter I have seen so far in TSW. This would be probably worse at high speed on ECML. Stutter happens every 6 to 7 seconds
Sounds should have been recorded from the actual train. If I am correct the OBB licence must have cost DTG a considerable amount. Surely access to the train for the purpose of sound recording should have been negotiated as a condition of paying for the licence?
Real life circumstances can prevent that. Renting a train for a full day or two while they have to run a timetable is very disruptive. Also a bunch of other things that could affect it
Hold up.. what a silly comment. If your new flag ship game does not have correct sounds for an add on included in the new base game, why wouldn't you play it safe and utilise something where it can be the authentic sounds for said locomotive? Release a DLC with improvised sounds, not your latest route/ Loco add-on...
Pretty sure its been mentioned that DTG had a recording session with Union Pacific here in the states for Sherman Hill, that got canceled cause of COVID, so they reused the prime mover sounds from another loco and got horn sounds from a person.
I dont think JT made the Pacer, I read somewhere (on this forum I think) that DTG made the Pacer a long time ago and it was intended for a release a long time ago however it never came up. Please correct me if I'm wrong but that's what I've read.
Whilst it is very disappointing the sounds aren't the correct ones, although personally I wouldn't know the difference, for me it doesn't mean the route should not be made. The route itself looks very impressive and interesting, the Talent 1 is very well modelled, I love the look of the mountains, for me the incorrect sounds of the Talent 1 are a minor point. I think some of those on this thread almost demanding the route does not get made for this reason should remember they are not speaking for everyone. Also calling them lazy, is insulting and shows a lack of thought. They have explained why the sounds are incorrect you either accept it or you don't.
I would like to grab that and make some annotations about making sound for trains (or generally). First, to ascribe laziness to someone without knowing the details, is just rude and disrespectful. Please respect the peoples work, even if you are not happy with the results. Back to the sound making. I often read here that the sounds should be synthesised when no sources available. So, what is sound that comes out of a synthesizer? How is it generated? Ho can we generate train sound out of nothing? I am dealing with synthesizers since 1993 or so. Nearly 30 years now for sure. As they are really nice things to play with, they can't do the impossible. Not the days back nor today. A synth uses sources too to generate sound. These often/mostly coming from the so called oscillators. An oscillator generates a tone out of a source (VCOs/DCOs out of voltage, Wavetable or Granular (what both uses waves as sources), or maybe more Math based technique (VA, Resonator). Without a source, there comes nothing out of the signal path (Mixer-Filter-Envelopes-Effects and what not else). No In, no Out (except maybe self oscillating filters, but they don't make train sounds for sure). With a standard synthesizer you can't make any train sound besides some base humming sound from transformers or such. For the Talents and other trains with inverter tech we need massively complex sound constructs. A synth with a thousand or more oscillators might be able to do that in a very synthetic way (the result would not sound like the real thing). There is not such a synth for reasons. It's cheaper then to buy the train to make recordings. The best approach to make train sound happen out of bad sources, is granular synthesis. But you need a source. No source input, no output, again. Granular synthesis is able to stop the forwarding of a source and plays it on a specific position as an type of loop. It slices the source in lots of little pieces (Grains) of a few ms length and layers them together again with some complex magic. With this tech you generally can record a trains acceleration as usual and then stop it at any point to loop it. The basis of what train sound needs to be dynamically controlled as in a TS, a loop. But this has bad downsides. It tends to sound like music, not like a machine. These granular synths are made for creating fascinating moving sound collages and other weird stuff that can be musically played with notes (in a scale at least, but don't need to). So, why not using the tech in TSW? Well. I tried it. UE4 4.2x comes with a granular synth plugin to use. I used a well recorded sound source from a bogie of the BR420. Made some spaghetti stuff in UE4 and fiddled with the settings. Solds out to nothing at least. It's not possible with the UE4 granular synth. Why? Settings! The parameter set is simply to small and offers not the things we need. The ranges are to tight (for Grain size, Grain length, Fade curve and curve types (no own curve types allowed)). So, the next step is to use a hardware granular synth (VST Instruments are also usable like UVI Falcon or MACH5 that are using the IRCAM Granular Algorithm ... plugins like Serum,Padshop,Pigments and such are not the best solution for that). That means a Waldorf Quantum or Iridium (the latter is not the best grab tbh.) and maybe for trying a GR1 from TastyChips. Of course, no grantee to get something out of these boxes. But the workflow is faster, the tech is better, more quality. It's worth a try but a try with a cost of nearly 5000€ or more (depends on how much different machines to try). But tbh. at least i don't think that these things can do the magic better than the others. A good example of using granular tech is the AP stuff from TSC and some of the TSW trains. But, if you have a ear on it more precisely, you can hear the musical basis of these sounds. It's well made, no question, but the sources were outstanding good and so are the results. The sources we would have for a Talent 1 would give nothing out of the granular. So, we are left to the usual things. Opening a wave editor (i'm using Steinberg Wavelab mostly) and try to make loops the old fashion way. That needs clean sources from a nice length of recordings. It should not swell up/down (what it generally does when recording a train, or it needs the train to maintain speed AND maintaining the power applied for a minute or so at certain speeds to get a set of loops from it). If the sources are not that nice, it starts to get ridiculously complicated and time consuming to get a loop from it. Often the time is not provided. We need to get fast results to stay economically. And again, the sources we would have for a Talent 1 are not capable to get loops out of them that would sound any good. It's better to use the Talent2 sounds because it is a sound that works at least (made by the same approach, swelling sound lopped by a nice simple technique .. but needs long and good recordings). Making sound for a train, that acts dynamically as needed, is hard work. Often not achievable at all. What also often stops the idea of making a specific train. Some might remember the VT98 story. Besides the physics problems in TSC, the sound was a main aspect to not bother with it a long time. It's still around, but i fear to make sound for it in TSW, because of all of YOU. So, no VT98 i would say.
Indeed. While we might be disappointed we are not getting the correct sounds, to accuse developers of being lazy is just rotten. Having watched more of the full stream and listened to the sounds my concerns have been lessened. Only thing I would say the cab volume level needs to be beefed up for entertainment value (maybe there should be a sound slider for this). Now Maik Goltz you know most of us would love a VT98 - the DB equivalent of the Pacer, so please do not be disheartened by those who choose to express themselves in a disagreeable manner.
Still pointing towards AI for train sounds. I never thought it was possible to recreate synthesised vocals like AI does. When I'm hearing all the gear Maik Goltz is mentioning I don't even understand how you would make engine sounds with them. Probably because I never tried but when I am making music and try to create the sounds I want I'm not pressured by time constraints. You really need to like noise to wanting to try to perfect an engine sound. I can understand you don't want to recreate an engine sound with a synthesizer. I think with AI you could build a tool that after a long time of training will give you the engine sounds you want in one click.
Vocals are easy for AI, dynamic machine sound will not happen for a long time. Especially not for such special sound needs like trains are. The results a AI would generate might be useful for a movie to use in the mix, but not for a simulator.
Don't train manufacturers like Siemens develop such AI for other uses? I think I read somewhere Daihatsu is working on AI for monitoring car engine sound.
Having seen the stream last night I saw and heard nothing that will stop me buying it assuming it has a reasonable smooth release. Just wanted to say Maik Goltz - your insight into TSW development is always really interesting to read even if it’s sometimes not what I want to hear (or can understand sometimes). I love to read what you and other developers such as Joe (the timetable guy) have to say.
but it gives you an output. The data is like a sheet of music but a sheet of music doesn't create audio as well.
It's the same as with knowing the frequencies and voltages of the inverters of a train or a loco. With the data you can't reproduce the audio of it. The sound is not generated by the inverters. It comes from the resonances of motor housings, the vibrations of bogies and chassis and all that. Pure inverter frequencies sounding odd and wrong. There are some examples on youtube from some European trains. And there are the Japanese guys that rebuild it with actual async motors, some wood and cardboxes. They use also such recorded data but using also a secret tech (not a chance to get onto that stuff).
Sounds awesome such installation. It's indeed that secret tech I was hoping for would be available. Audio is mostly geared towards vocals and instruments less towards sounds we like to avoid in audio recordings like engines.
It was also stated that during the pandemic, Deutsche Bahn closed its depots and non-public facilities to all non-employees.
I think it was that DTG planned to make the pacer for TVL but never made it, instead we had the Class 101. But I dont think they actually made it though
Correct, I believe. Matt once said during a livestream that they had the Pacer on some sort of internal schedule (implying it would have been a loco DLC for Tees Valley), but one day it just got removed before ever being seriously worked on and Matt was quite sad. Who knows… maybe a planned recording trip didn’t happen, they had to allocate resources to other content etc.
I've wanted to ask this question for a long time. Why can one person do 1000 times better than a developer(https://www.trainsimcommunity.com/mods/c3-train-sim-world/c20-sounds/i3743-class-700-sound-mod)? After all, DTG had access to Class 700 sound files, which did not stop them from creating this mess instead of propper sounds.
They only had access to the ac sounds from ap. Ap never gave them the dc sounds. Dtg was only able to record sounds from a static 700, not a moving one. So rather than having no sounds at all, a 450 was recorded.
The 700 is a slightly different issue, the main problem being the relative levels, the cab is just too quiet. There is also a bit of a loop evident too. TBH I haven’t driven it all that much though have the mod installed (which probably won’t work with TSW4) that makes it at least tolerable.
Take it from someone who who has travelled abourd many class 700's, Dtg did NOT make a mess of it, sure the audio levels need tweaking in places but overall it still sounds very much like the real thing.
Mixed sounds? from what i can hear from the live stream its exactly the same sounds as the BR 442 Talent 2. Despite all the great work you've all done from on the quality of the route to I'm really looking forward to the more changes of this game, it's just sound's are so important too. Soon when Nahverkehr Dresden -Riesa in the deluxe edition comes out there will be another double on = DB BR 442 audio. Only two regrets, one of them is the lack of St. Margrethen station. and the other is Audio from the Talent 2 in a Talent 1 Just the same as Skyhook Games ever did with the BR 187 along with 185/2 audio (a long time ago of course). most similarities can be heard at this time (32:25 ) In Train Simulator classic is a video on YouTube of nice realistic traction sound of the Talent 1. It would be nice if it could be made in that direction if necessary as others have said a MIX of the class 377
The earlier mentioned Japanese people. The stuff you see there is a secret. At least the software and special inverter hardware they use, not the motors. And its not even sounding near the real thing.
During the preview we hardly had any outside up close views for us to hear the sounds of the 4024. But when it was the Vectron it was a different,they got up and close for us to hear. Honestly not sure what’s more concerning night lighting which from what I saw is the same and BCC, or the sounds. Overall is a great route that I’m looking foward to, bu the lighting and sounds are concerning….
I have to highly disagree with this. Sound is extremely important, that's why even racing games put so much work into it. I will raise a comparison - TSW Br110 and SR EU07. The locos represent a relatively similar internal architecture. Sounds in EU07 have so many more details, each click when changing taps, rumble on switches, and loud cooling. It is so much more enjoyable to listen to it compared to Br110. Sound is crucial.
It's old and rare source material that was used and there as nothing else to get. Same for lots of other trains. Seems i am not capable enough to make sound for modern games. As soon as the editor is out, i will give up making audio for the trains. It's not even any fun at all since a whole while. There are more talented people out there to do that. If someone really talented and with access to trains in Germany wants to work with me on the trains i plan to do (not the DTG ones, my own DLCs). please contact me.
In my opinion the sounds of the 110 are great, any trial to make it more perfect is going in the "99% effort to get closer another 1%" area. This is beyond pareto and even for me working in the railway industry difficult to hear any difference.
You already heard them in the preview video that was uploaded before the stream. Nothing changed in terms of sound
When I heard, that an austrian route is coming, I was like shut up and take my money. After the first video of the 4024 it changed. The route seems nice, but I can't accept the copy and paste sounds. Not for the price DTG asks for. Sure there will be a sale in the future. So long I stay with TSW 3.
Yes I did. And that’s when I gave my initial opinion on the sound and that was reinforced after the hearing them on the stream mostly inside the cab ,the outside view was mostly avoided. Since they said the sounds are final I know I must learn to live with them. That doesn’t mean I’m happy with it and doesn’t mean I can’t make comments about it. Same goes with the lighting
You mean other than in the UK having a work from home unless a key worker policy for much of 2020 and international travel all but banned for 18 months and even if for essential then having quarantine restrictions on return. Thats just the UK policy and German & US will have had their own restrictions in place as well to varying degrees. If you need proof of the above I suggest you try google. Or are you saying DTG providing you with a train with prototype sound recordings is essential and that they should have put their staff and the staff of their licenses at risk just so you could drive with realistic sounds?
