i7 Why should I i7? - Video Editing.

[EvelynX]

Every day I hear "But the i7 is better for video editing" but I never bothered to ask just how much better?

Do any of you fine folks have links to some tangible evidence of just how much better an i7 is over it's i5 equivalent for video editing and rendering. How much time is someone going to save exactly?
What I'd like to see is a video edited with the same CPU at the same clock speed but with hyper-threading both enabled and disabled so I can see just how much time exactly is saved.

Can anyone help me find that information?

[desertcomputer]

More threads and more cache which video software likes a lot. 

[EvelynX]

2 minutes ago, MrUnknownEMC said:

More threads and more cache which video software likes a lot. 

That's great and all but you must have only read the title. Can you help with my question?

[Guest]

6 minutes ago, ChocoROID said:

That's great and all but you must have only read the title. Can you help with my question?

Go look at cinebench scores done with i5's and i7's

What your asking for has literally never been done. Its also common sense, More things working at an objective, the faster it'll get done

[desertcomputer]

2 minutes ago, ChocoROID said:

That's great and all but you must have only read the title. Can you help with my question?

Well, the question you ask is not reasonable.. the reason you get i7 is because of additional threads. If you do benchmark of i7 with only 4 threads (No hyper-threading) what is point. Clock speeds aren't make equal, also faster ram will impact on too, the i7 will only helps with render time by around 40% faster really depend what effects and video format. It a common sense ... more threads = more work can be done at same time which result in faster render times. 

[Lays]

The i7 is something like 20-30% faster, sometimes a bit more I believe.

 

So that means if you had to render a video that would take 10 hours on the 4690k, it may only take 7 hours on a 4790k.

 

 

Here's someone rendering something with both CPU's, the 4790k was 28% faster, that 28% is a big difference the longer it takes to render something means more time saved.

 

[Blueprint]

Cinebench score/render time is a great example (as said by @Ronnie76). 

 

http://www.maxon.net/products/cinebench/overview.html

[EvelynX]

Just now, Ronnie76 said:

Go look at cinebench scores done with i5's and i7's

What your asking for has literally never been done. Its also common sense, More things working at an objective, the faster it'll get done

Of course it's better and of course it's faster, but nowhere am I able to find the exact amount of time (probably as a percentage) how much better it is? How do you measure the value if you don't know that. "Oh this renders the video in 18 minutes at 4 cores but 12 with seven."
I'm finding it hard to believe that nobody has tested this.
If it's a good 20% that's great. If its 5% then that's a problem? Surely someone has tangible evidence, but nobody can tell me exactly how much faster it is.

Honestly I'm surprised LTT hasn't done this.

[EvelynX]

3 minutes ago, MrUnknownEMC said:

Well, the question you ask is not reasonable.. the reason you get i7 is because of additional threads. If you do benchmark of i7 with only 4 threads (No hyper-threading) what is point. Clock speeds aren't make equal, also faster ram will impact on too, the i7 will only helps with render time by around 40% faster really depend what effects and video format. It a common sense ... more threads = more work can be done at same time which result in faster render times. 

You miss the point of the question.

[EvelynX]

3 minutes ago, Lays said:

The i7 is something like 20-30% faster, sometimes a bit more I believe.

 

So that means if you had to render a video that would take 10 hours on the 4690k, it may only take 7 hours on a 4790k.

 

 

Here's someone rendering something with both CPU's, the 4790k was 28% faster, that 28% is a big difference the longer it takes to render something means more time saved.

 

This is exactly what I'm looking for. Tangible evidence. Do you have the source? I'd like to see if core speed impacted these results.

[Lays]

2 minutes ago, ChocoROID said:

Of course it's better and of course it's faster, but nowhere am I able to find the exact amount of time (probably as a percentage) how much better it is? How do you measure the value if you don't know that. "Oh this renders the video in 18 minutes at 4 cores but 12 with seven."
I'm finding it hard to believe that nobody has tested this.
If it's a good 20% that's great. If its 5% then that's a problem? Surely someone has tangible evidence, but nobody can tell me exactly how much faster it is.

Honestly I'm surprised LTT hasn't done this.

It's just from a video on YT, here's a link

 

 

 

 

 

I'm sure if you look hard enough with google searches you can find lot's more evidence / tangible stuffs.  

 

For the most part, if you just look at any synthetic benchmark that can properly utilize every thread, then compare both scores and get the %, the % increase in speed with the 4790k will basically scale perfectly into any other program that can utilize the threads. 

Basically the same goes for clock speed, it's nearly linear for increase in clock speed vs increase in score.   IE: 10% overclock usually will net you nearly 10% of extra performance.

[desertcomputer]

3 minutes ago, ChocoROID said:

Of course it's better and of course it's faster, but nowhere am I able to find the exact amount of time (probably as a percentage) how much better it is? How do you measure the value if you don't know that. "Oh this renders the video in 18 minutes at 4 cores but 12 with seven."
I'm finding it hard to believe that nobody has tested this.
If it's a good 20% that's great. If its 5% then that's a problem? Surely someone has tangible evidence, but nobody can tell me exactly how much faster it is.

Honestly I'm surprised LTT hasn't done this.

Because it really depends ... how much effects and video format and read speed of the drives. Why do you need evidence ... it common sense. You can just JUST a google search it.  

[EvelynX]

Just now, valdyrgramr said:

http://cpu.userbenchmark.com/Compare/Intel-Core-i7-4790K-vs-Intel-Core-i5-4690K/2384vs2432  Click dis.  It's will break down why.

Reading only the title.

[EvelynX]

6 minutes ago, Lays said:

It's just from a video on YT, here's a link

 

 

 

 

I'm sure if you look hard enough with google searches you can find lot's more evidence / tangible stuffs.  

 

For the most part, if you just look at any synthetic benchmark that can properly utilize every thread, then compare both scores and get the %, the % increase in speed with the 4790k will basically scale perfectly into any other program that can utilize the threads. 

Basically the same goes for clock speed, it's nearly linear for increase in clock speed vs increase in score.   IE: 10% overclock usually will net you nearly 10% of extra performance.

Thanks buddy, you've been the most helpful so far.
The i7 here has a higher core speed so it doesn't show how much time is saved based solely on the extra cores but for the most part it indicates the time value of the better CPU.
What I really wanted to know was the time improvement because "Faster" is meaningless without a number on there in minutes.

If its 20-30% faster and thats how much time I'll save on average, then I can actually work out if the extra money spent on an i7 is worth the gains. For video editing with the informaiton you've given me, I'd say yes!
If it was only like 5% faster though, I wouldnt be too happy.

[EvelynX]

1 minute ago, valdyrgramr said:

 

Because programs that do video editing take advantage of single thread, multicore/thread, and other speeds.  On top of that the i7 has hyperthreading, so you're using 8 vs 4 threads.  The 8 threads won't get as congested in said tasks, so it can help get them done faster.  The 4690k can do it too, but the i7 will get it done faster.  If time is money to you then get the 4790k, but if you don't care about time then go with the 4690k.  Another option is a cheap xeon called the E3-1231.

Thanks but I know all that.
I was looking to find out how much actual time is saved with 8 over 4. Time quantified as a number in seconds or minutes.

Faster is a meaningless statement when working out price/performance.
But you try telling that to a Comcast salesperson! "Fast" is how they advertise!

[ClobberXD]

1 minute ago, ChocoROID said:

Thanks but I know all that.
I was looking to find out how much actual time is saved with 8 over 4. Time quantified as a number in seconds or minutes.

Faster is a meaningless statement when working out price/performance.
But you try telling that to a Comcast salesperson! "Fast" is how they advertise!

So you want benchmarks? Sorry that I can't produce one, but they can be obtained from Google - Cinebench is a good benchmarking utility, and so is Hand-brake, if you want to compare the i5 with the i7 in terms of video-editing.

[EvelynX]

1 minute ago, anandgeforce said:

So you want benchmarks? Sorry that I can't produce one, but they can be obtained from Google - Cinebench is a good benchmarking utility, and so is Hand-brake, if you want to compare the i5 with the i7 in terms of video-editing.

Will these results tell me how many seconds or minutes I'd be saving buying i7 instead of i5?

[ClobberXD]

5 hours ago, ChocoROID said:

Will these results tell me how many seconds or minutes I'd be saving buying i7 instead of i5?

Cinebench produces results as a 'score', while Handbrake will tell you the exact number of seconds (or minutes) it took...