Ok, blahcomp222, you've already got some good advice - I'm going to amplify.
I have an ASRock Extreme6 (overkill for your interests). I've had ASRock boards going back years now. One $60 board lasted 7 years doing 24/7 without maybe a day off once or twice a year - and it had no solid caps. It gave up, finally, after a GTA V binge my son put it through before I noticed lint had completely clogged the HSF. An $80 ASRock ran 24/7 for 5 years, and is still in service (replaced the $60 version when it quit).
Some will knock ASRock, but then some have unkind words for ASUS, Gigabyte, MSI....but the fact is ASRock has the value/performance market. It's tough to find a board to match the price with similar features, and I can't find real, tangible reasons to favor any of these boards. I was once an MSI fan (circa 2002), but at this point they're last on my list (they have some great exceptions).
The Extreme3 suggested (though I haven't had one of those) is a great choice.
Locked Xeons, or even the non-K versions of 1150 chips would be in line with your goals. Many of the free video products/encoders seem to assume 4 cores/threads, so HT may SEEM inapplicable, but...
You CAN run two instances at once with many such applications, where HT has some benefit to you. Aside from that, HT gives you availability for other tasks while video processes continue in the background. Unless the budget gets really tight, favor an HT version.
RAM can usually be purchased at the lowest price for the parameters you require without much regard for a specific brand/model. There are a few "don't buy" options, like white label (no brand) and certain "too good to be real" bargains. If 16 Gbytes is under $115 US as of Feb, 2015, it's probably not real. It could be the wrong RAM, like overstock registered ECC RAM that's not applicable to most 1150 boards.
You'll get a lot of warnings about brands, but there are only a few RAM chip makers in the world, and most sticks are fair at stock speeds - it's fair to stick with known brand names, but don't fall into paying high dollar for brand names because they're suggested. I've had AData and Transcend sticks, without any decorative heat spreaders, run for years (like 10 years) without incident.
I can't argue against the Ripjaws suggested thus far, but I loathe tall heatsinks. I actually avoid heatspreaders on RAM entirely if the price is better (and I've never had a reason to regret that), but my last purchase was G.Skill Ares series, with short, red heatspreaders that look great (and were on sale with rebate at the time) - excellent results.
There are RAM favorites, especially for overclockers, but for stock speeds you just need to focus on a few things (aside from being DDR3 for a DDR3 board, non ECC)...
First, prefer 1.5v. Higher performance RAM may be rated for higher than 1.5v, and that usually works out ok, but the standard spec is 1.5v and it means the speed ratings are....shall we say more native? Anyway, prefer 1.5v over higher voltages.
Anything over 1600 is overkill for your requirements - but sometimes deals come up where 1866 or even 2400 is on sale. As of today, it seems that 1333 and 1600 are within $1 (or $0). Both will work fine, and you will not notice any real time performance difference, and you may not even notice benchmark differences except in very rare cases.
There is a catch, and it's the other parameter that matters most in RAM. CAS (sometimes called CL). It's the first number in the timing sequence. Consider these two specs:
1333 9-9-9-24
1600 11-11-11-28
These are RAM timings for two different sticks of 8Gbyte chips (1). The "double" clock speed is the first number (you probably know), CAS (or CL) is the first number in the collection of 4 timings (all 2 digit or less).
It's better to have lower CAS. This represents the time it takes for RAM to point to a completely new address. The other timings are of limited importance (except to RAM overclockers). However, if any RAM parameter is going to matter at all in your selection, it's CAS.
Here's where it gets confusing: CAS is related to the RAM clock speed.
In reality, CAS is a REAL TIME delay. That is, there is a WALL CLOCK amount of time the circuits require in order to do their work no matter what clock speed is selected, and they can't easily go faster. If you buy RAM rated for, say, 2400 on a special sale, the CAS listed is relative to the 2400 speed.
Let's say you found a sale on 2400 11-11-11-28 RAM
If you ran the 2400 RAM at 1600, which it will do fine, the CAS would be set lower than 11 (likely 9, could be 8 - I'll check in a moment). This is not overclocking, it happens to be correct stock because this is a REAL TIME variable, relative to the RAM's operating speed.
In other words, comparing the CAS speed of, say, 1333 RAM (commonly 9) to the CAS speed of, say, 2400 RAM (commonly 11) does not mean the CAS of 11 is actually slower...because it's RELATIVE to the clock speed.
Here's how:
2400 RAM would be clocked at 1200 (it's doubled)...the CAS is set as a count of those 1200 ticks...so let's say that's as CAS of 11 (typical for 2400 RAM).
At 1200 million ticks per second, one tick takes 0.00000000083333333333 of a second, or 0.833 nanoseconds. (it's 1/1200 million).
11 of those is about 9.1667 nanoseconds of real time.
Now, let's say you ran that RAM at 1600. That's clocked at 800Mhz, which means each tick is 0.00000000125000000000 of a second, or about 1.25 nanoseconds.
8 * 1.25 = 10....or 10 nanoseconds of real time.
Notice that's actually longer than the 9.1667 calculated for the 2400 @ CAS of 11?
This means that 2400 RAM of CAS 11 would commonly run at CAS of 8 if you operated at 1600 Mhz.
You'll notice that most 1600 RAM is sold with CAS of 9, and 11 is also common (much slower, longer CAS than this 2400 would be set).
Since CAS is an integer, you can't dial in really tight timings, but as long as you honor the rated real time delay (let it take what real time the RAM requires), running higher speed RAM at a slower setting can offer, frequently, the benefit of tighter CAS timings WITHOUT overclocking anything.
1333 and 1600 are close matches, but usually 1600 CAS 9 can run at CAS 8 when clocked at 1333.
It also happens to be that 1600 with CAS 11 can't usually run at CAS 9 when clocked at 1333...it usually has to be 10, because if you repeat the math above you'd find that for 1600, CAS of 11 relates to 13.75 ns, where as at 1333 CAS of 9 would be 13.5 ... a little faster than the rating, so it has to be a 10.
Why do I state that?
Well, if you pay for 1600 CAS 11 (which is a bargain speed), but ran it at 1333 CAS of 10 to be complaint, the actually time of your delay would be 15 ns.
That RAM could run at 13.75ns, but if you're underclocking it to 1333 your only choices are CAS of 11 = 16.5ns, CAS of 10 = 15 ns, or the CAS of 9 which is too fast at 13.5 ns.
This math applies to 1866 you might run at 1600.
Let's say you find a bargain 1866 of CAS 9 on sale. That RAM can run CAS at 9.64630 ns, which is much faster than the 1600 CAS 11 we just examined.
Let's say you run that at 1600 (you're not overclocking your board remember, 1866 is an overclock speed).
So, at 1600 your options are....CAS 9 = 11.25 ns, CAS 8 = 10 ns, CAS 7 = 8.75 ns....and CAS 7 is too fast for that 1866 RAM, so CAS 9 = 10 ns is your best STOCK SPEED choice.
When I search for RAM, I setup a spreadsheet that does this math, look for the bargains, and judge how I might actually run that RAM. I don't overclock RAM (it's a waste to bother really)...but I do want the best CAS performance for my purchase, and because CAS is an integer, there are odd alignments sometimes, and they can be deceiving.
To realize that a 1600 CAS 9 might actually operate faster than a downclocked 1866 CAS 11 may automatically make sense to you.
To realize that a 2400 CAS 11 downclocked to 1600 would run FASTER for you than a stock 1600 CAS 9 may not automatically make sense to you, but it's true.
RAM prices, sales, deals...they move so fast that making a recommendation is tough (you could miss a deal).
Especially for a target where RAM is not to be overclocked, knowing this bit of math can help you find a real bargain...or avoid overpaying for something that offers little or no real benefit.
Real world CAS differences of more than 1.25 nanoseconds is tangible performance gains.
......and once again I've gone on until everyone is asleep, right?
right....
