Color in 2600 games seem dull on the 7800?

[kyle]

I just got another 7800 mine has been broken for a while now and I've been just playing my 2600. Well I have noticed before that the color in 2600 games seems to just not look as bright on the 7800 as they do the 2600. Well after only playing 2600 for over a year and hooking up my new 7800 last night I can say for sure that when playing a 2600 game on a 7800 the color just isn't as bright. Does anyone know why?

[Tubular Gearhead]

The video circuit on the 7800 mixes the 2600 signal with the 7800 signal. The result is a compromise in signal strength and color in both modes. If the 2600 signal is adjusted (by changing some component values), there will be unacceptable wavy lines when playing 7800 games. It is unfortunately just the way it is unless you modify the video circuit. I was not happy with this compromise so I changed the circuit. See below for one of my prior posts...

 

"I see no reason to not believe that Best developed this modification from something that was done years ago. There are several different ways to build a simple video amplification circuit for the 7800; I developed mine prior to reading about the "easier video mod". One of my goals was to have crystal clear video in both 2600 and 7800 modes so the solution I came up with was to get rid of the entire rf circuit, separate the chroma signals (stock they are blended and the result is a compromise in color and image quality) and use the old channel select switch to choose between 2600 and 7800 modes. I was able to tailor the modification to have bright, non washed out colors in both 2600 and 7800 modes with individual brightness pots for each (the stock luma ladder values have been optimized for the one stage amplification circuit).

 

Anyway, the point is there are many ways to get composite video out of the 7800, and I think it is very possible that some engineer modified the console Best found. "

 

I didn't draw out a schematic ( ) , or I would post it.

Good luck.

Chris

[kool kitty89]

That's odd, I wouldn't have thought a simple switch (be it automatic or manual) would have cost much, and if the compromise was THAT significant it's certainly an odd thing to leave out.

 

That's at the composite video level, right, feeding into the RF modulator?

[Tubular Gearhead]

That's odd, I wouldn't have thought a simple switch (be it automatic or manual) would have cost much, and if the compromise was THAT significant it's certainly an odd thing to leave out.

 

That's at the composite video level, right, feeding into the RF modulator?

 

That's correct. It's in the circuit which feeds the rf modulator. Basically, I think the specifications were for somebody to be able to just plug in any cartridge, be it 2600 or 7800, and have the machine play it w/o the person having to slide a switch putting the video circuit in the correct mode. As you said, it could have been done automatically (digitally, MARIA takes over when a 7800 cartridge is plugged in, but the analog signals stay blended), but I guess it was cheaper to just mix all the signals, and work with the values and levels until the compromise was acceptable (although as pointed out, it is noticeable and I don't consider it acceptable).

 

Remember, the 7800 is basically a complete, independent 2600 as well as the 7800 Maria system on one motherboard. MARIA has separate Chroma, luma and Sync signals, and so does TIA. They are all blended, and basically what I did was separate them, remove several components (rf modulator, audio choke- since you don't need to blend audio with video anymore, etc), and restructure the circuit so that the signals are stronger. It really isn't that hard- if you look at the schematics in the AtariAge archives you will see how it is laid out (bottom right hand corner of schematic).

 

But now I have to move a switch every time I take out a 7800 cartridge and put in a 2600 cartridge- Not really a big deal if you ask me.

Edited January 20, 2011 by Tubular Gearhead

[toiletunes]

What a relief! I've been playing 7800 in 2600 mode for a long time, but I just got a 2600 and I couldn't figure out how to turn down the oclors.

[Tubular Gearhead]

What a relief! I've been playing 7800 in 2600 mode for a long time, but I just got a 2600 and I couldn't figure out how to turn down the oclors.

 

[kool kitty89]

That's correct. It's in the circuit which feeds the rf modulator. Basically, I think the specifications were for somebody to be able to just plug in any cartridge, be it 2600 or 7800, and have the machine play it w/o the person having to slide a switch putting the video circuit in the correct mode. As you said, it could have been done automatically (digitally, MARIA takes over when a 7800 cartridge is plugged in, but the analog signals stay blended), but I guess it was cheaper to just mix all the signals, and work with the values and levels until the compromise was acceptable (although as pointed out, it is noticeable and I don't consider it acceptable).

Couldn't they have used a digital switch (triggered either by booting in 7800 mode or a physical signal on the cart slot) that toggled between MARIA and TIA composite output -only ever connecting one analog video signal at a time? (in the case of a dedicated cart signal, that would eat up one of the few added cart pins... unless maybe they could hack one of the existing VCS pins for such: assuming shield-ground was properly grounded on every single VCS game released, they could pull it high to toggle to MARIA video and at the normal grounded state for VCS -though given the system already specifically differentiated between modes on the software side, such a mechanism might not have been useful -unless it was cheaper/simpler than using MARIA/7800 booting to toggle the video)

 

 

Hell, even with a manual switch, they could have left it convenient for the layman (or lazy people who didn't care about color/video quality) and had a mixed setting by default with an on/on/on switch toggling from TIA to both to MARIA with the factory default set to mixed to be fairly foolproof and still avoid any added logic for automatic switching.

Edited January 20, 2011 by kool kitty89

[Tubular Gearhead]

That's correct. It's in the circuit which feeds the rf modulator. Basically, I think the specifications were for somebody to be able to just plug in any cartridge, be it 2600 or 7800, and have the machine play it w/o the person having to slide a switch putting the video circuit in the correct mode. As you said, it could have been done automatically (digitally, MARIA takes over when a 7800 cartridge is plugged in, but the analog signals stay blended), but I guess it was cheaper to just mix all the signals, and work with the values and levels until the compromise was acceptable (although as pointed out, it is noticeable and I don't consider it acceptable).

Couldn't they have used a digital switch (triggered either by booting in 7800 mode or a physical signal on the cart slot) that toggled between MARIA and TIA composite output -only ever connecting one analog video signal at a time? (in the case of a dedicated cart signal, that would eat up one of the few added cart pins... unless maybe they could hack one of the existing VCS pins for such: assuming shield-ground was properly grounded on every single VCS game released, they could pull it high to toggle to MARIA video and at the normal grounded state for VCS -though given the system already specifically differentiated between modes on the software side, such a mechanism might not have been useful -unless it was cheaper/simpler than using MARIA/7800 booting to toggle the video)

 

 

Hell, even with a manual switch, they could have left it convenient for the layman (or lazy people who didn't care about color/video quality) and had a mixed setting by default with an on/on/on switch toggling from TIA to both to MARIA with the factory default set to mixed to be fairly foolproof and still avoid any added logic for automatic switching.

 

Yes they could have. I don't know why they didn't- usual 7800 suspects I guess- money, motherboard space, added complexity, etc...

 

It has been a few years since I did this modification, so I don't remember exactly what I did, but I might revisit it again and see if I can't change it for automatic switching. I am sure there are several ways to do it.

 

Anyway, it is almost shocking to see how good the composite video looks in both 2600 and 7800 modes when all the signals are kept separate. It is amazingly bright, colorful and clear... Go figure.

[+Random Terrain]

It has been a few years since I did this modification, so I don't remember exactly what I did, but I might revisit it again and see if I can't change it for automatic switching. I am sure there are several ways to do it.

 

Anyway, it is almost shocking to see how good the composite video looks in both 2600 and 7800 modes when all the signals are kept separate. It is amazingly bright, colorful and clear... Go figure.

Maybe then I would start to love my Atari 7800 and not be so dissapointed. And once the Atari 7800 gets its own version of batari Basic, I'll love it even more.

[+GroovyBee]

though given the system already specifically differentiated between modes on the software side, such a mechanism might not have been useful -unless it was cheaper/simpler than using MARIA/7800 booting to toggle the video)

 

It differentiates in hardware too. When Pin 5 on the LS174 (U11 on the NTSC schematic) is low then MARIA is disabled. When pin 2 of U11 goes high the video mode is "locked".

 

If MARIA is disabled you can enable your own 2600 AV mod. After the 7800's BIOS has determined the cart type and its found to be a 2600 one then the video mode is "locked" so that 2600 games can't change it back.

[Tubular Gearhead]

though given the system already specifically differentiated between modes on the software side, such a mechanism might not have been useful -unless it was cheaper/simpler than using MARIA/7800 booting to toggle the video)

 

It differentiates in hardware too. When Pin 5 on the LS174 (U11 on the NTSC schematic) is low then MARIA is disabled. When pin 2 of U11 goes high the video mode is "locked".

 

If MARIA is disabled you can enable your own 2600 AV mod. After the 7800's BIOS has determined the cart type and its found to be a 2600 one then the video mode is "locked" so that 2600 games can't change it back.

 

Ok, so this got me thinking and looking at the schematic again to try and remember what exactly I did. If you look at U3 (logic gate) you will see that the 2600 and 7800 Luma and Sync signals are indeed switched. The Chroma signals, however, are not. Specifically, the value of R17 has been chosen high so the 2600 signal will not override the 7800 signal causing interference when MARIA is enabled. So, at the very least, R16 and R17 should be replaced with pots and switched before combining @ L1. This would be a good spot for another logic gate (I think that is where I used the old channel switch). If I remember right, I also reduced the values of R11, 12, 13, & 15 by about half, preserving the spread of the Luma logic ladder (is that what it is called?), but making it easier to amplify w/o adding additional stages. I am pretty sure I just used only one transistor. I changed the value of R10, but I am not sure if I added a diode to the Sync or not- I think it was better w/o one. I don't think it was necessary to amplify the Sync. I will try and take some photos of the mod and draw up a schematic at some later time.

[kool kitty89]

Another thing Atari/GCC could have done is omit internal RF entirely in favor of a plain AV port and external modulator box (like several home computers of the time) and put BOTH video signals on the port (perhaps use the A8 pinout with a DIN-5 and swap luma for an 2nd composite signal), not only moving the problem to an external solution but cutting down on the cramped 7800 board space (especially the early revisions with the riser board over the modulator), maybe even enough to make them reconsider adding POKEY. (though using a single 8x8k SRAM also could have been smart forward thinking -higher component costs in the short run, but less board space and much greater savings in the long run as well as the utility of added RAM)

 

If they had done that, you'd still either need a manual external switch for 2600/7800 modes or logic to throw a switch between the 2. (in the latter case you'd probably have to bump the AV port to a 7-pin DIN if you were going to put the switching mechanism on the RF box side... but that really wouldn't make much sense as it would take similar -if not simpler- circuitry to do it all internally and again have only 1 composite output...) The manual switch option would probably be fine for most users and the cheapest route, though if they released AV cables, you'd also need a switch on that.

 

Actually, using an external RF box like that could have allowed for an automatic RF switch mechanism that switched to game mode when fed 5V (also powering the modulator) sort of like the 5200 but without the funky multiplexed power/RF combo. (more like what Sega did on the model 2 Genesis and model 1 in Japan)

In that case, you'd definitely want the 2600/7800 switch at the base of the cable on the 7800 side as you could then bury the actual RF box/switch mechanism behind the TV and not worry about being able to switch it. (though putting the 2600/7800 switch on an even more accessible location on the console itself would have been more convenient and would then allow later models to add an auto switch mechanism to displace the manual switch -plus with 1 composite like you also simplify cable design and allow the exact pinout for the 8-bit monitor cable sans luma)

[prostx23]

Hi,

 

I never really noticed a problem with the video in 2600 mode (perhaps I hgaven't been paying close enough attention), but I always thought the colors in 7800 mode always looked very muted and never understood why. I guess I know now.

 

Mike

[0078265317]

What a relief! I've been playing 7800 in 2600 mode for a long time, but I just got a 2600 and I couldn't figure out how to turn down the oclors.

 

 

2600 graphics are bad enough to begin with I don't know how you could get any worse.

[kool kitty89]

What a relief! I've been playing 7800 in 2600 mode for a long time, but I just got a 2600 and I couldn't figure out how to turn down the oclors.

 

 

2600 graphics are bad enough to begin with I don't know how you could get any worse.

Try playing through noisy RF. (especially with a TV with poor or no fine tuning/automatic fine tuning)

[Atarifever]

What a relief! I've been playing 7800 in 2600 mode for a long time, but I just got a 2600 and I couldn't figure out how to turn down the oclors.

 

 

2600 graphics are bad enough to begin with I don't know how you could get any worse.

2600 graphics are often fantastic. Kaboom, Solaris, Radar Lock, H.E.R.O., Pitfall 2,etc are all beautiful games.

[psquare75]

My 7800 gathers dust as the colors look so bland with the 2600 games.

[0078265317]

What a relief! I've been playing 7800 in 2600 mode for a long time, but I just got a 2600 and I couldn't figure out how to turn down the oclors.

 

 

2600 graphics are bad enough to begin with I don't know how you could get any worse.

2600 graphics are often fantastic. Kaboom, Solaris, Radar Lock, H.E.R.O., Pitfall 2,etc are all beautiful games.

 

Not really. Pitfall 2 looks much beter on the atari 5200 than on the 2600 version.

[Phredreeke]

IMO the 5200 version barely looks better than the 2600 version (sure, there's some more detail in the trees...)

[kool kitty89]

IMO the 5200 version barely looks better than the 2600 version (sure, there's some more detail in the trees...)

More detail in the background in general. (caves, ground, trees, etc) OTOH it also seems cut-down with no use of DLIs for the (much easily managed) equivalent to the VCS's CPU timed faster effects (per scanline color reloading), rather odd. (the C64 version was more detailed in some respects but also had the ugly black background used)

 

 

I think the sounds a bit weaker compared to the VCS version though except the noisy mixing necessary for Crane's hack to work (I think it feeds 4-bit PCM that the CPU plays on one of TIA' channels via volume modulation -not sure of any other way you could mix audio digitally other than 1-bit PWM which would be FAR more CPU intensive).

I think the DPC added more channels than POKEY (or the CV for that matter -or C64) could do in hardware (seems like it added somethign fairly close to the SN76489's 3 square wave plus 1 noise channels plus 1 TIA channel for SFX). A bit odd that they used a square wave for the A8 jump sound rather than periodic noise (which seems to have been simulated in software by the CV using PWM -also on the Tandy/PCJr though that might be through the PC speaker for 5 total channels like the VCS, and the C64 used the hardware pulse wave I think)

The A8 version does have a nice clean and bright sound to it compared to the others (including the CV).

Edited January 25, 2011 by kool kitty89